You are on page 1of 37

Almost 40 Years of Airframe Noise Research What did we achieve?

Werner Dobrzynski DLR, Institute of Aerodynamics and Flow Technology 14th Aeroacoustics Conference, 5-7 May 2008 / Vancouver
Folie 1 > AIAA/CEAS-2008-AFN > Dobrzynski


Problem definition and historical overview Noise sources and mechanisms Noise reduction technologies Summary of achievements and future needs

This informative talk focuses on experimental and applied research in airframe noise and is an attempt to subsume related activities worldwide without claiming to be complete.

Folie 2 > AIAA/CEAS-2008-AFN > Dobrzynski

EU: Visions 2020 Goals / US: NASA-AST/QAT

European Visions 2020: Reduce noise impact by one half per operation relative to 2000 technology. 0
-5 Noise Reduction -10 -15 -20 -25 -30 -35 -40 1990 2000 US-Goals EU-Visions 2020 2010 Time 2020 2030

United States Noise Reduction Goals: Reduce perceived noise impact of future aircraft by one half relative to 1997 technology within 10 years (AST and QAT Program) and by three quarters (-20 dB) within 25 Years!

Note: Reduction by one half is -10 dB subjectively, corresponds to - 90% in sound power
Folie 3 > AIAA/CEAS-2008-AFN > Dobrzynski

Why are we Interested in Airframe Noise ?

Airframe noise is generated through the interaction of turbulent flows with solid bodies, e.g. landing gears and lifting surfaces. With the introduction of high-bypass ratio engines around 1970 (for fuel saving !) a significant reduction in jet noise was achieved As a result airframe noise became relevant in the approach phase, where engines are throttled down Substantial research efforts into airframe noise started after 1970
B707 A380

1th generation engines

2nd generation engines

Folie 4 > AIAA/CEAS-2008-AFN > Dobrzynski

Overall Aircraft Noise Ranking at Approach Aircraft Noise Source Breakdown:

APPROACH (long range aircraft) APPROACH (short range aircraft)

Source: Airbus

Folie 5 > AIAA/CEAS-2008-AFN > Dobrzynski

Concise Survey of Airframe Noise Research

Airframe noise was first identified as a lower aircraft noise barrier in the early 70ies Main objectives were to quantify airframe noise levels through dedicated flyover noise tests ranging from gliders to the Galaxy aircraft Fink developed a first Airframe noise prediction method essentially based on this data (1977) Already at that time fundamental noise reduction concepts were invented and wind tunnel tested on generic models Between about 1980 to 1990 the interest in airframe noise temporarily decreased (due to the first fuel crisis) From 1990 up to now airframe noise reduction again is a matter of prime interest
Folie 6 > AIAA/CEAS-2008-AFN > Dobrzynski

History of Noise Reduction Technologies

Based on theoretical work in aeroacoustics in the 60ies, already from 1975 on most of the basic noise reduction technologies were invented:
Trailing edge/ slat noise porous edge extensions (Bohn, 1976) perforated edge extensions and edge serrations (Grosche, 1979 and Fink, et. al 1980) porous leading edge inserts (Fink, et. al 1980) porous edge replacements (Fink, et. al 1980, and later Revell, et. al 1997) TE-brushes slat cove cover; -filler; -liner side-edge fences, -brushes, -porous inserts moldline technology (for hinged flaps) application of streamlined fairings

Flap noise Side-edge noise

More current developments are:

Trailing edge/ slat noise Flap side-edge noise Landing gear noise

Folie 7 > AIAA/CEAS-2008-AFN > Dobrzynski

Major Sources of Airframe Noise

Folie 8 > AIAA/CEAS-2008-AFN > Dobrzynski

Parasitic Airframe Noise Sources Landing Gears Tone noise from pin-holes :
NLG tow-bar pin holes
Tones originate from flow excited resonances in different pin holes.

Torque link pin holes

Parasitic tone noise can easily be avoided by pin-hole covers, but such means are not popular due to water condensation problems.
Folie 9 > AIAA/CEAS-2008-AFN > Dobrzynski

Parasitic Airframe Noise Sources Wing Systems

SoundPressure LevelLevel Sound Pressure [dB]

Tone noise from fuel vents or anti-ice vents Broadband noise from slat track cut-outs Broadband noise from flap side-edge cavities
A320 Slat-track cut-outs: Fuel overpressure vent

Wing Cavity Tones Fan-Tones

10 dB

Baseline Low noise

v = 105 m/s f = 14 Hz
Baseline All cavities sealed Cruise Configuration 0 500 1000 1500 2000 Frequency Frequency[Hz] [Hz]

A320 Flap edge cavity

A-weighted Level

Example: MD-11 A-weighted Level 10 dB 10 dB

Baseline Sealed cavities


Baseline Clean Flap-edge







Folie 10 > AIAA/CEAS-2008-AFN > Dobrzynski

Second Order Airframe Noise Sources

Airframe noise sources of minor importance:
Wing tip noise (seldom identified in source location plots due to other prominent sources) Noise from infinitely extended surfaces (e.g. fuselage) is expected to be more than 10 dB below current high-lift noise levels (i.e. below clean aircraft trailing edge noise), this is boundary layer noise including roughness effects noise from surface vibration (low radiation efficiency due to impedance mismatch between wall material and air) Free wake turbulence (quadrupole noise is insignificant for M < 0.2) Tonal vortex shedding noise from smooth circular landing gear struts (seldom observed for high flight Reynolds numbers)

Folie 11 > AIAA/CEAS-2008-AFN > Dobrzynski

Landing Gear Noise Problem Definition

More than 10 dB reduction demonstrated for not practical complete aerodynamic fairing Future efforts needed to realise noise reduction of similar order of magnitude under practical constraints 1/3-oct. Band Level First full scale LG test in 1995:
A320 Main Gear in German-Dutch Wind Tunnel

10 dB
Baseline Low Noise Medium



Folie 12 > AIAA/CEAS-2008-AFN > Dobrzynski

Landing Gear Noise - Source Mechanisms

Landing gears represent a 3D cluster of noise sources:
Broadband noise from turbulent vortex shedding off struts Turbulent wake-flow / solid body interaction between components Broadband noise from gear bay free shear layer interaction with downstream bay rim

Landing gear noise

scales on a Strouhal number basis increases with velocity to the 6th power (dipole type source) and features an almost omnidirectional radiation characteristic
Folie 13 > AIAA/CEAS-2008-AFN > Dobrzynski

Landing Gear Low Noise Design - Constraints

Limitation of runway loads defines number of wheels and spacing Gear locations defined by lateral stability and rotation before lift-off (local velocity at gear position for under wing installation) Brake cooling (fairings which would delay cooling increase the turn around time on airport)

Free fall requirement (MLG leg door can not be used as spoiler) Tyre burst (location and redundancy of dressings)

Weight Effect on airframe (minimum bay size for stowing) Systems complexity (articulated components) Maintenance (fairings obstruct quick inspection, contamination)
Folie 14 > AIAA/CEAS-2008-AFN > Dobrzynski

Landing Gear Noise Baseline WT-Studies

Boeing Low Speed Aeroacoustic Facility

Baseline full scale or close to full scale LG tests

B777 1/16 scale MLG

Folie 15 > AIAA/CEAS-2008-AFN > Dobrzynski

Landing Gear Noise Add-on Fairing Design

Farfield wall mounted microphones

A340 fairings manufactured to airworthiness requirements

Flight testing of landing gear fairings: - 2 EPNdB achieved on total A340 landing gear source noise

Herkes, et. al. / AIAA 2006-2720

Airbus MLG Fairings NLG NLG and MLG

Folie 16 > AIAA/CEAS-2008-AFN > Dobrzynski

Landing Gear Noise Elastic Fairings

2 dB


scale B777 gear

Cloth fairings promised an additional 2 dB in noise reduction compared to solid fairings.

Source: Ravetta, et. al. / AIAA 2007-3466

Folie 17 > AIAA/CEAS-2008-AFN > Dobrzynski

Landing Gear Noise Advanced Low Noise Design

Iterative design process for low noise gears based on experimental experience with support from CFD calculations: DNW-LLF Test:
Design with support from CFD computation Full scale mock-up wind tunnel test Test result: 5 to 7 dB(A) noise reduction In-flight prediction: - 4.1 EPNdB on LG-level
Lm - 60log(v/vref) (dB) 80
= A340 original NLG90 RAIN add-on fairings SILENCER WP2.3.2

5 dB

60 1

10 Strouhal Number fms/v


Folie 18 > AIAA/CEAS-2008-AFN > Dobrzynski

Landing Gear Noise - Prediction Tools

Empirical and Semi Empirical Tools:
Fink (NASA ANNOP) (empirical tool scaled from early DLR scale model data) DLR (empirical tool re full scale wind tunnel data by Dobrzynski) ISVR (semi empirical tool by Smith) Boeing (semi empirical tool by Guo) Still only limited design to noise capability

Source: M. Smith / ISVR

Semi empirical tool concepts:

Superposition of noise contribution from individual components Account for small details through complexity or dressing factor (governs high frequency levels)

CFD/CAA approaches with design to noise capabilities:

LES/ DES calculations and FW-H for simplified gear geometries only (not really convincing results yet)
Folie 19 > AIAA/CEAS-2008-AFN > Dobrzynski

High-Lift Devices Noise Problem Definition

1/10.6 scaled A340 model in DNW-LLF 8 m by 6 m open test section

Challenges in HLD noise testing in comparison with LG experiments:

Testing of complete wing systems mostly performed at small scale (i.e. 1/4 up to 1/11) due to lack of sufficiently large anechoic facilities. Accordingly results suffer from: Reynolds number effects (tone noise)
1/3.8 scaled 777 semi span in NASA Ames 40 ft x 80 ft closed test section

Components can not always be manufactured to scale (trailing edge thickness, track design) Basic investigations in 2D (i.e. no sweep) need validation in 3D (cross flow) Wing downwash causes inaccurate aerodynamic conditions in open test sections for measurement of farfield noise directivity
Folie 20 > AIAA/CEAS-2008-AFN > Dobrzynski

High-Lift Devices Noise Sources

For the same speed (!) noise from high-lift devices in landing configuration is about 10 dB higher compared to the same aircraft in clean conf. (cruise). Major HLD noise sources are:
Slotted slats Slat tracks (oblique to inflow) Slat horn Flap side-edges Flap tracks
t Sla
Total wing noise


Source: Oerlemans / NLR

Folie 21 > AIAA/CEAS-2008-AFN > Dobrzynski

High-Lift Slotted Slats - Noise Mechanisms

Source: Choudhari, / NASA Langley

Analytical and CFD/ CAA studies provided insight in major slat noise mechanisms. PIV Measurements:

Vorticity field

Potential source mechanisms: Free shear layer vortex flow reattachment Trailing edge (bluntness tone noise only relevant at model scale !) Unsteadiness of vortex core
Folie 22 > AIAA/CEAS-2008-AFN > Dobrzynski

High-Lift Slat Noise Sources of Tone Noise


Low frequency tones

1/7 scale model

High frequency tones

Tone phenomena for scale model testing: Low frequency tone effects due to cavity resonances (Rossiter modes ?) (to be reduced by tripping on pressure side)

Source: Oerlemans / NLR

High frequency tone effects due to Tollmien-Schlichting instability (to be avoided by tripping on suction side)
Folie 23 > AIAA/CEAS-2008-AFN > Dobrzynski

High-Lift Devices - Slat Noise Characteristics

Directivity from WT tests:

Farfield noise test results at a 1/7.5 scaled A320 model and a full scale A320 wing section indicate that
Slat noise dominates the total high lift noise spectrum

However, details of source mechanisms and related local geometric and flow parameters can only be determined through dedicated CAA computations.

Slat noise scales with the 4.5th power of flight speed and wetted span (SF) vs a Strouhal number, based on slat dimension Slat noise slightly decreases with increased angle-of-attack (low -range) Slat noise radiation is highest in the rear arc.
Folie 24 > AIAA/CEAS-2008-AFN > Dobrzynski

High-Lift Devices - Slat Noise - Effect 1/10.6 scaled A340 Model in DNW-LLF: = m + = 103

Landing range


Slat noise dominated

Only little effect on slat noise of moderate landing angles-of-attack, i.e. < 10 Significant noise increase for > 13
Folie 25 > AIAA/CEAS-2008-AFN > Dobrzynski

Tools Development Computational Aeroacoustics

Sound Pressure Level / dB

Stochastic broadband approach Low cost CAA

Narrow Band Spectra

CFD: Steady RANS

Time averaged turbulent flow problem


CAA based Sound Propagation (time domain)

Acoustic farfield

Stationary turbulent properties:

Unsteady vortex/entropy sound sources

Simulation (PIANO) Exp. (EADS-IW AWB)


RPM(*): Stochastic model for turbulent fluctuations (time domain)

(*) RPM: Random ParticleMesh method, in press Comp. & Fluids

5 5 kHz

10 10

Local length scale / turb. kinetic energy / anisotropy

APE = Acoustic Pertubation Equation TKE = Turbulence kinetic energy RPM = Random particle mesh TKE from RPM Reconstruction
Folie 26 > AIAA/CEAS-2008-AFN > Dobrzynski

TKE RANS target

Turbulence source terms

High-Lift Devices Low Noise Design Constraints

Maximum lift (determines landing speed) 10 % less CLmax is about 5.4 % increase in landing speed = 1.4 dB noise increase !! Sufficient lift for moderate angles-of-attack to prevent tail-strike for take-off

Reliability (low noise treatments must not affect performance if not operational) No sudden lift/ moment changes through activation of control device

Weight Structural constraints (slat tracks affect front spar position, etc.) Systems complexity (e.g. bleed air for flow control, etc.) Maintenance (contamination, icing of noise red. treatments)
Folie 27 > AIAA/CEAS-2008-AFN > Dobrzynski

High-Lift Devices Slat Noise Reduction

3 dB noise reduction through slat-cove-cover

With extended blade seal

Source: Khorrami, / NASA Langley

Folie 28 > AIAA/CEAS-2008-AFN > Dobrzynski

High-Lift Devices Slat Noise Reduction

Slat noise reduction through cove filler:
Predicted with CFD/CAA Validated by experiments of different research groups (i.e. DLR, EADS-IW, NASA and JAXA), But noise reduction found to be extremely sensitive re filler contour. Inboard

Mid span

Source: Horne, / NASA Ames
Folie 29 > AIAA/CEAS-2008-AFN > Dobrzynski

High-Lift Devices Slat Noise Reduction

A320 Full Scale and Model Scale Tests: NACA0012 Airfoil Test:
1 mm 1 mm TE-comb 0.4mm 0.4 mm
s s

TE comb

40 mm 40 mm u

45 mm
TE thickness:

slotted TE s = 0.1 mm 0.4 mm

0.15 mm
Reference (with flap-edge brush) With additional slat TE-brush

Noise Reduction, dB


flexible TE comb stiff TE comb flexible TE-comb stiff TE-comb slotted TE

s < 0.1 mm s = 0.1 mm u = 40 m/s 50 m/s 60 m/s

Slat trailing edge brushes are effective for noise reduction but installation is sensitive re CLmax degradation (!) In contrast slat trailing edge serrations showed a slight noise increase





0 = 1 mm
Folie 30 > AIAA/CEAS-2008-AFN > Dobrzynski

High-Lift Devices Low Noise Slat Design

Noise reduction to be achieved through
reduced gab size and increased overlap reduced slat deflection angle with adapted flap deflection for constant aerod. performance

Source noise reduction (array):

VLCS | Run 1005 | dpt3 | Ma=0.1 | Alpha=4.5 3E Ref | Run 1098 | dpt3 | Ma=0.1 | Alpha=6.0

F15, M=0.2, Re=12x106 5.0 4.5 4.0

VLCS Reference

1/3-Octave Band Level [dB]




4 dB

3.0 2.5 CL 2.0 1.5


reference VLCS-CL,max VLCS-L/Dmax

1.0 0.5
60 2000 4000 6000 8000

1/3-Octave Band [Hz]

0.0 -10.00






Folie 31 > AIAA/CEAS-2008-AFN > Dobrzynski

High-Lift Devices Flap Side-edge Noise Reduction

Brush- or porous edge:

Source: Choudhari, et. al., NASA Langley

Full wing noise reduction

Clean Flap-edge Brush Edge, high density, 0.055

Side-edge Fences:

Brush Edge, low density, 0.055

Fences provide some 2-3 dB source noise reduction


Local source noise reduction

Moldline Techno. for hinged or Fowler flap

Baffled Flap Side-Edge

Source: NASA
Folie 32 > AIAA/CEAS-2008-AFN > Dobrzynski

High-Lift Devices Spoiler Noise

Spoilers can be applied to enable steep noise abatement approach procedures and to contribute to wake vortex alleviation.
Baseline fm = 12.5 kHz TED = 60 + Spoiler 6 = 20 dB

Outboard spoiler deflection:

Local slat noise reduction due to spoiler deflection

Source: Oerlemans / NLR

Conventional spoilers are sources of low frequency noise, but spoiler source noise mechanisms are still not known in detail (edge noise, wake flap interaction noise?) Note: Spoiler deflection affects wing circulation and in turn slat noise !!

Folie 33 > AIAA/CEAS-2008-AFN > Dobrzynski

Interaction Noise Sources Gear wake/ Flap

AWB Artefacts


X [m]

Noise Reduction [dB]

Parametric wind tunnel study on 1/10 scaled wing-gear model: Wake flow measurements Interaction noise measurements Result: up to 10 dB source noise increase at low frequencies 2 dB high-lift devices noise reduction through porous flap LE liner

0,5 0,0 -0,5 -1,0 -1,5 -2,0 -2,5 -3,0 1,000 Frequency [kHz)
Folie 34 > AIAA/CEAS-2008-AFN > Dobrzynski

Flap LE liner


To answer the question: What did we achieve?

Detailed insight in noise source mechanisms but still no sound knowledge of all noise generation parameters for complex 3D sources Computational Aeroacoustics (CAA) methods have developed rapidly and promise to enable a design-to-noise of airframe components in the future Landing gear source noise reduction of up to 5 EPNdB has been demonstrated for future low noise gear designs High-lift devices source noise reduction is still limited to less than 1 EPNdB (through add-on means for conventional slats) and often suffers from a degradation in maximum lift Low noise slat designs (for constant aerod. performance !) were identified but need validation in 3D Flap side-edge low noise modifications were identified and validated
Folie 35 > AIAA/CEAS-2008-AFN > Dobrzynski

Status and Future Needs

Status of Airframe noise reduction: Airframe noise reduction achievements still need further improvement to cope with the European Vision 2020 and the even more stringent US noise reduction goals While LG noise reduction is on a good track high-lift devices noise reduction needs more attention in particular with respect to maintain aerodynamic performance Future needs: Focussed efforts to develop CAA tools for 3D application in the industrial design chain Flow separation control technologies must be developed for both gear structures and slat-less high-lift devices In long terms new aircraft configurations are needed, which feature short landing gears and enhanced lift capabilities for reduced approach speed.
Folie 36 > AIAA/CEAS-2008-AFN > Dobrzynski

Thank you for your attention !

Folie 37 > AIAA/CEAS-2008-AFN > Dobrzynski